Drive - RAID Technology Overview

Hierachy Storage

About

Redundant array of independent disks (RAID) is the technology of grouping several physical drives in a computer into one or more logical drives.

Each logical drive appears to the operating system as a single drive.

RAID is all about:

  • protecting the data (RAID 5 and 6 are most efficient fault tolerant)
  • performance (RAID 1 and 10 are highest performance)

The ideal RAID solution consists of balancing trade-offs between:

For instance, if data availability is your primary concern, the lower performance of a RAID 6 may be worth it, especially for applications such as web servers or video-on-demand (VOD) systems where performance is generally indistinguishable.

Performance

This grouping technique greatly enhances logical-drive capacity and performance beyond the physical limitations of a single physical drive.

When you group multiple physical drives into a logical drive, the Server RAID controller can transfer data in parallel from the multiple drives.

This parallel transfer yields data- transfer rates that are many times higher than with non-grouped drives.

This increased speed makes the system better able to meet the throughput (the amount of data processed in a given amount of time) or productivity needs of the multiple-user network environment.

The ability to respond to multiple data requests provides not only an increase in throughput, but also a decrease in response time. The combination of parallel transfers and simultaneous responses to multiple requests enables disk drives to provide a high level of performance in network environments.

RAID controller / bus

Choosing a RAID controller currently means choosing between a PCIe bus interface and a PCI-X interface; and it’s important to make the selection that will suit your needs. PCI-X has the advantage of being backwardly compatible with the older PCI interface. But it is parallel and half-duplex bidirectional, and the bus runs only as fast as the slowest device. PCIe is the new forward-looking standard, and is intended to cope with the performance and scalability demands for at least the next decade. PCIe has the advantages of being serial, full-duplex bidirectional, and devices are able to independently negotiate the bus speed.

Stripe-Unit Size

Definition

With RAID technology, data is striped across a group of physical drives.

This data-distribution scheme complements the way the operating system requests data.

The granularity at which data is stored on one drive of the logical drive before subsequent data is stored on the next drive of the logical drive is called the stripe-unit size.

You can set the stripe-unit size to 16, 32, 64, 128, 256 (the default), 512 or 1024 KB. You can maximize the performance of your Server RAID controller by setting the stripe unit size to a value that is close to the size of the system I/O requests.

Note: The maximum supported stripe size for RAID 6 and RAID 60 is dependent on the number of drives in the array. In general, the more drives in the array the smaller the maximum supported stripe size.

Example

For example, performance in transaction-based environments, which typically involve large blocks of data, might be optimal when the stripe-unit size is set to 64 KB or 128 KB.

However, performance in file and print environments, which typically involve multiple small blocks of data, might be optimal when the stripe-unit size is set to 16 KB. The collection of stripe units, from the first drive of the logical drive to the last drive of the logical drive, is called a stripe.

Selecting a RAID Level and Tuning Performance

Disk logical drives are used to improve:

  • performance
  • and reliability.

The amount of improvement depends on the application programs that you run on the server and the RAID levels that you assign to the logical drives.

Each RAID level provides different levels of:

  • fault-tolerance (data redundancy),
  • utilization of physical drive capacity,
  • and read and write performance.

In addition, the RAID levels differ in regard to the minimum and maximum number of physical drives that are supported. When selecting a RAID level for your system, consider the following factors.

RAID level Data redundancy Physical drive capacity utilization Read performance Write performance Built-in spare drive Min. number of drives Max. number of drives
Simple Volume No 100% Superior Superior No 1 1
RAID level-0 No 100% Superior Superior No 2 16
RAID level-1 Yes 50% Very high Very high No 2 2
RAID level-1E^ (Enhanced) Yes 50% Very high Very high No 3 16
RAID level-5^ Yes 67% to 94% Superior High No 3 16
RAID level-5EE** Yes ^^ 50% to 88% Superior High Yes 4 16
RAID level-6^ Yes 50% to 88% Very high Fair No 4 16
RAID level-10 Yes 50% Very high Very high No 4 16
RAID level-50** Yes 67% to 94% Superior High No 6 128
RAID level-60** Yes 50% to 88% Very high Fair No 8 128
Spanned Volume No 100% Superior Superior No 2 48
RAID Volume No 50% to 100% Very high to Superior * Fair to Superior * No 2 48

* Depends upon underlying RAID level. ** Available with ServeRAID-8i only. ^ Not available with ServeRAID-8k-l. ^^ RAID level-5EE is not redundant while it is compressing.

Physical drive utilization, read performance, and write performance depend on the number of drives in the logical drive.

Generally, the more drives in the logical drive, the better the performance.

RAID controllers can also support the following additional RAID levels :

  • Simple Volume - a single disk drive or segment; not redundant
  • Spanned Volume - two or more disk drives or segments with the same or different capacity, connected end-to-end. A spanned volume offers no redundancy or performance advantage over a single drive
  • RAID Volume - two or more logical drives with the same RAID level, connected end-to-end. The logical drives may have the same or different capacity and are not striped together; they may be redundant, depending on the RAID level

Note: RAID volumes can be created from RAID level-0, RAID level-1, or RAID level-5 members, but RAID levels cannot be mixed within the same RAID volume.

RAID Level-0

RAID level-0 stripes the data across all the drives in the logical drive. This offers substantial speed enhancement but provides no data redundancy. RAID level-0 provides the largest storage capacity of the RAID levels that are offered, because no room is taken for redundant data or data-parity storage. RAID level-0 requires a minimum of two drives and, depending upon the level of firmware and the stripe-unit size, supports a maximum of 8 or 16 drives. The following illustration shows an example of a RAID level-0 logical drive.

Start with two physical drives Two Physical Drives
Create a logical drive using two physical drives Logical Drive Using Two Physical Drives
The data is striped across the drives, creating blocks.
Notice that the data is striped across all the drives in the logical drive,
but no redundant data is stored.
Data Is Striped Across The Drives

A physical drive failure within the logical drive results in loss of data in the logical drive assigned RAID level-0, but only in that logical drive. Logical drives assigned RAID level-1, level-1E, level-5, level-5EE, or level-6 do not lose data. Note: RAID level-5EE is not redundant while it is compressing, so if a drive failure occurs during this state, data loss is possible. When you replace a failed drive, the ServeRAID controller can rebuild all the RAID level-1, level-1E, level-5, level-5EE, and level-6 logical drives automatically onto the replacement physical drive. However, any data stored in a failed RAID level-0 logical drive is lost. Although the risk of data loss is present, you might want to assign RAID level-0 to a logical drive to take advantage of the speed this RAID level offers. You can use this logical drive to store data that you back up each day and can re-create easily. You also might want to use a RAID level-0 logical drive when you require maximum capacity. RAID level-0 offers the following advantages and disadvantages.

Advantages Disadvantages
• Substantial speed enhancement
• Maximum utilization of physical drive storage capacity, because no room is taken for redundant data or data-parity storage
No data redundancy, resulting in data loss in the event that a physical drive fails

RAID Level-1

RAID level-1 uses data mirroring. Two physical drives are combined into a logical drive, and data is striped across the logical drive. The first half of a stripe is the original data; the second half of a stripe is a mirror (that is, a copy) of the data, but it is written to the other drive in the RAID level-1 logical drive. RAID level-1 provides data redundancy and high levels of performance, but the storage capacity is diminished. Because the data is mirrored, the capacity of the logical drive when assigned RAID level-1 is 50% of the drive capacity. RAID level-1 requires two physical drives. The following illustration shows an example of a RAID level-1 logical drive.

With RAID level-1, if one of the physical drives fails, the controller switches read and write requests to the remaining functional drive in the RAID level-1 logical drive. RAID level-1 offers the following advantages and disadvantages.

Start with two physical drives Two Physical Drives
Create a logical drive using the two physical drives Logical Drive Using Two Physical Drives
The data is striped across the drives.
Notice that the data on the drive on the right is a copy.
Raid1   Data Is Striped Across The Drives

of the data on the drive on the left.

Advantages Disadvantages
• 100% data redundancy
• High performance
Allows only 50% of the physical drive storage capacity to be used

RAID Level-1 Enhanced

RAID level-1 Enhanced (RAID level-1E) combines mirroring and data striping. This RAID level stripes data and copies of the data across all of the drives in the logical drive. As with the standard RAID level-1, the data is mirrored, and the capacity of the logical drive is 50% of the drive capacity. Note: RAID level-1E is not supported on ServeRAID-8k-l. RAID level-1E has a similar profile to RAID level-1; it provides data redundancy and high levels of performance, but the storage capacity is diminished. However, RAID level-1E allows a larger number of physical drives to be used. RAID level-1E requires a minimum of three drives and, depending upon the level of firmware and the stripe-unit size, supports a maximum of 8 or 16 drives. The following illustration is an example of a RAID level-1E logical drive.

With RAID level-1E, if one of the physical drives fails, the ServeRAID controller switches read and write requests to the remaining functional drives in the RAID level- 1E logical drive. RAID level-1E offers the following advantages and disadvantages:

Start with three physical drives Three Physical Drives
Create a logical drive using the physical drives Logical Drive Using Three Physical Drives
The data is striped across the drives, creating blocks Raid1e   Data Is Striped Across The Drives

Notice that the stripe labeled
∗ is the data stripe and the stripe labeled
∗∗ is the copy of the preceding data stripe.

Also, notice that each block on the mirror stripe is shifted one drive.

Advantages Disadvantages
• 100% data redundancy
• High performance
Allows only 50% of the physical drive storage capacity to be used

RAID Level-5

RAID level-5 stripes data and parity across all drives in the logical drive.

RAID level-5 offers both data protection and increased throughput. When you assign RAID level-5 to a logical drive, the capacity is reduced by the capacity of one drive (for data-parity storage). RAID level-5 gives you higher capacity than RAID level-1, but RAID level-1 offers better performance.

RAID level-5 requires a minimum of three drives and, depending upon the level of firmware and the stripe-unit size, supports a maximum of 8 or 16 drives.

The following illustration is an example of a RAID level-5 logical drive.

Start with four physical drives Four Physical Drives
Create a logical drive using three of the physical drives, leaving the fourth as a hot-spare drive Raid5   Logical Drive Using Three Of The Physical Drives
The data is striped across the drives, creating blocks.
Notice that the storage of the data parity (denoted by ∗) also is striped,
and it shifts from drive to drive.
A parity block (∗) contains a representation of the data from the other blocks in the same stripe.
For example, the parity block in the first stripe contains data representation of blocks 1 and 2.
Raid5   Data Is Striped Across The Drives
If a physical drive fails in the logical drive, the data from the failed physical drive is reconstructed onto the hot-spare drive. Raid5   Data From The Failed Physical Drive Is Reconstructed Onto The Hot Spare Drive

RAID level-5 offers the following advantages and disadvantages.

Advantages Disadvantages
• 100% data protection
• Offers more physical drive storage capacity than RAID level-1 or level-1E
Lower performance than RAID level-1 and level-1E

Documentation / Reference





Discover More
Windows Disk
Data Storage - Boot Sector Startup Processes

Two sectors are critical to start a computer on Windows: the master boot record (MBR), the first sector of a hard disk and the boot sector, which resides at sector 1 of each volume. These sectors...
Card Puncher Data Processing
Data Storage - Data Striping (I/O parallelism)

In computer data storage, data striping is the technique of segmenting logically sequential data, such as a file, in a way that accesses of sequential segments are made to different physical storage devices....
Card Puncher Data Processing
Data Storage and Manipulation (Computer Architecture)

Computers are all manipulating Data. The efficiency of a computer is a function of: how fast it can manipulate the data (CPU speed) and how fast it can retrieve and store that data (CPU cache,...
Hierachy Storage
Drive - (Storage|Disk) Array

A disk array is a disk storage system which contains multiple disk drives. It is differentiated from a disk enclosure, in that an array has cache memory and advanced functionality, like: RAID JBOD....
Hierachy Storage
Drive - Disk (hard disk) controller

The disk controller (or “hard disk controller”) is an circuit embedded in a hard disk and is the equivalent of a miniaturized computer, as it contains a small CPU and its own memory to obtain and transfer...
Data Path From Disk To Cpu
IO - Data Path / Balanced System

The data paths is the path composed of all hardware components that are needed to get the data from: a storage device (generally disk drive or network drive) to the CPU It is important to understand...
Card Puncher Data Processing
IO - IOPS (Input/Output Operations Per Second)

IOPS (Input/Output Operations Per Second, pronounced eye-ops) is a common performance measurement used to benchmark computer storage devices like: hard disk drives (HDD) solid state drives (SSD),...
Card Puncher Data Processing
Oracle Database - ASM (Automatic Storage Management)

Starting with Oracle Database 10g (10.1.0.3 or later), Automatic Storage Management is a high-performance storage management solution. For Oracle Database files, it simplifies the management of a dynamic...
Card Puncher Data Processing
Oracle Database - ASM - Disk group

A disk group is a set of disk devices that Automatic Storage Management manages as a single unit. Each disk device can be: an individual physical disk, a multiple disk device, such as a RAID storage...
Card Puncher Data Processing
Oracle Database - File System

The file system is a storage option for the database files during the installation of a database. Oracle Database Configuration Assistant creates the database files in a directory on a file system mounted...



Share this page:
Follow us:
Task Runner